The field of the present invention relates generally to the calibration of signal sources; and, more particularly to a method of calibrating a signal sourcing device by selecting the signal sources in the device in a manner which tends to cancel out errors in the signal sources.
Many devices contain signal sources and select signals generated from the signal sources. However, signal sources do not generate precise signals. As a result, the overall accuracy of the device is compromised by the accuracy of the signal sources.
An example of a signal sourcing device is a digital-to-analog converter (DAC). High speed DACs are implemented by a current steering architecture. As the value of the digital code input to the DAC increases, the number of current sources steered to the output also increases. The overall accuracy of the DAC is limited by the accuracy of the current sources. A uniform error in absolute accuracy results in a gain error which may be tolerated at times. The small errors between individual current sources, due to the matching limitations, result in errors in the output of the DAC as the digital codes inputted to the DAC change. These errors in the current sources limit the linearity and accuracy of the DAC. As a result, engineers have attempted to reduce the error in the accuracy of the DAC by, for example, calibrating the current sources in the DAC.
Prior art methods of calibrating a DAC included various techniques. One technique includes the calibration of the current sources in the DAC by using external, high quality test equipment. For example, the DAC may have an array of current sources including extra current sources. A technician would use the high quality test equipment to measure the current output by each current source and select the most accurate current sources for use in the DAC. However, this approach cannot remember the proper current sources to select for the particular integrated circuit containing the DAC unless if a non-volatile memory is used such as laser trimming or fuses. Thus, another approach was to use the external high quality test equipment to determine the accuracy of each current source and then trim each current source to become more accurate. Trimming current sources or the components of current sources, is a costly and slow procedure. For example, an eleven bit DAC may use 2047 current sources and the cost and time to trim that many current sources is very substantial.
Yet another prior approach was to attempt to do on-chip calibration of the current sources. The calibration could occur at power on, or whenever requested, if implemented on an integrated circuit. One difficulty with an implementation in an integrated circuit is the difficulty in implementing a reference and calibration circuit with sufficient accuracy on an integrated circuit. The reason is that, for example, for an eleven bit DAC, the accuracy of the calibration circuit must be greater than eleven bits which is very hard to accomplish. On-chip calibration techniques for a DAC include rotating the selection of the current source for each cycle of the DAC in order to average the current sources so that the DAC effectively sees no large variations between the different current sources. Another on-chip calibration technique for a DAC is to measure the accuracy of each current source on the chip with an accurate analog-to-digital converter (ADC), a costly endeavor, and then to trim the current sources to be more accurate, another costly and time consuming process.
Thus, there is a need for a method to calibrate DACs in a cost and time effective manner by improving the accuracy of the current sources which in turn improves the overall accuracy of the DAC. There is also a need to create a more accurate DAC where the accuracy of the DAC can be repeatedly executed without having to recalibrate the DAC every time.
Further limitations and disadvantages of conventional systems will become apparent to one of skill in the art after reviewing the remainder of the present application with reference to the drawings.
Various separate aspects of the present invention can be found in a method of calibrating a device having a plurality of signal sources, each source generating a signal, by categorizing the signal sources into groups based on the amount of error in the signal generated by the signal sources and selecting the signal sources in the device from the groups in a manner which reduces errors in the selection process.
The signal sourcing device may be, for example, a DAC having a plurality of current sources, each current source generating a current. The current sources generate currents having a median value, where some of the currents have errors relative to the median value. The improved DAC categorizes the current sources into groups based on the amount of error in the current and selects the current sources from the groups in a manner which reduces the overall error in the selection process. Because the overall accuracy of a DAC is limited by the accuracy of the current sources, the improved DAC reduces the effect of errors between individual current sources on the output of the DAC. More specifically, the improved DAC selects a first current value, determines whether the current generated by each current source of the plurality of current sources is greater than, substantially equal to, or less than the first current value, determines whether the number of the current sources whose current is greater than the first current value is more than, less than or substantially equal to the number of the plurality of current sources whose current is less than the possible median current value, and adjusts the first current value to be closer to the median current. The improved DAC may repeat these steps until the number of the current sources whose current is greater than the first current value is equal or substantially equal to the number of the plurality of current sources whose current is less than the possible median current value. At this time, the first current will approximate the true median current value. The improved DAC may then categorize each current source into one of a plurality of groups based on the current generated by the current source, wherein each group covers a mutually exclusive range of currents. In one embodiment, each group of current sources has a complementary group of current sources. In another embodiment, there is a center group of current sources whose range of currents straddles the true median current and each of the remaining groups of current sources has a complementary group. Thus, by categorizing the current sources in the digital to analog converter into groups based on the amount of error in the current generated by the current sources, the improved DAC can select the current sources in the digital to analog converter from the groups in a manner which reduces errors in the conversion. For example, the DAC may select the current sources from complementary groups where current sources in complementary groups tend to have similar but opposite errors. This calibration method may be used to reduce errors in the current sources which have a linear error distribution, a Gaussian error distribution, a combination of linear and Gaussian error distribution, or other error distributions or their combinations.
The improved signal sourcing device may be used in any device having a plurality of signal sources such as a television, tuner, settop box, computer, computer component such as a display or monitor, vehicle, antenna, broadcast receiver, broadcast transmitter, wireless handset, modem, camera, or other imaging device.